GAIA: Geothermal Analytics and Intelligent Agent

TL;DR

Geothermal field development is highly multidisciplinary and data-intensive under tight deadlines. GAIA introduces an integrated agentic AI platform with three components—GAIA Agent for planning, GAIA DT for modular physics-based and surrogate modeling, and GAIA Chat for user interaction—built around a retrieval-augmented generation workflow. The system uses Gemma3-27B-IT as the main agent, a multi-modal embedding and vector store for knowledge retrieval, and a seismic-focused digital twin (ObsPy-based) with interactive visualizations, enabling automated phase picking, event location, magnitude estimation, and seismicity forecasting. Experiments demonstrate GAIA’s ability to automate seismic analytics across four use cases with an extensible roadmap toward domain-specific LLMs, MCP tool integration, surrogate modeling, and self-evolving multi-agent architectures. This work presents a pathway toward scalable, data-driven, and automated geothermal project workflows that can accelerate decision-making and operations in the field.

Abstract

Geothermal field development typically involves complex processes that require multi-disciplinary expertise in each process. Thus, decision-making often demands the integration of geological, geophysical, reservoir engineering, and operational data under tight time constraints. We present Geothermal Analytics and Intelligent Agent, or GAIA, an AI-based system for automation and assistance in geothermal field development. GAIA consists of three core components: GAIA Agent, GAIA Chat, and GAIA Digital Twin, or DT, which together constitute an agentic retrieval-augmented generation (RAG) workflow. Specifically, GAIA Agent, powered by a pre-trained large language model (LLM), designs and manages task pipelines by autonomously querying knowledge bases and orchestrating multi-step analyses. GAIA DT encapsulates classical and surrogate physics models, which, combined with built-in domain-specific subroutines and visualization tools, enable predictive modeling of geothermal systems. Lastly, GAIA Chat serves as a web-based interface for users, featuring a ChatGPT-like layout with additional functionalities such as interactive visualizations, parameter controls, and in-context document retrieval. To ensure GAIA's specialized capability for handling complex geothermal-related tasks, we curate a benchmark test set comprising various geothermal-related use cases, and we rigorously and continuously evaluate the system's performance. We envision GAIA as a pioneering step toward intelligent geothermal field development, capable of assisting human experts in decision-making, accelerating project workflows, and ultimately enabling automation of the development process.

Figures (6)

• Figure 1: GAIA Agentic RAG Workflow. Users interact with GAIA Chat interface by supplying prompts and supplementary data. The GAIA Main Agent will first think through the prompt and make a step-by-step plan to solve the given problem. If necessary, the system will fetch data from its knowledge base consisting of PDFs of papers/proceedings, tabular data, and/or external data sources (via web search and/or GDR). The collected data and the formulated workflow will be forwarded to GAIA DT for the actual processing through its set of tools. Finally, the Main Agent will formulate the final response based on the gathered information and multi-step analyses.
• Figure 2: GAIA Chat interface. The main window features a chatbox to type and send messages, a conversation window to display message history, a file upload button, and a search setting panel to control the number of returned search items from the knowledge base. The interactive figure output allows users to pan, zoom, and hover over the generated figures, and export them to PNG or PostScript (PS).
• Figure 3: Example of GAIA's seismic waveform phase picking workflow. The user uploads a seismic waveform file and requests phase picking (a). GAIA Agent will then analyze the waveform data, apply the selected phase picking method, and return the results in an interactive plot (b).
• Figure 4: Example of GAIA's event location estimation workflow. The user uploads a seismic waveform file and requests event magnitude estimation (a). GAIA Agent will then analyze the waveform data and apply the selected location estimation methods (b--c).
• Figure 5: Example of GAIA's event magnitude estimation workflow. The user uploads a seismic waveform file and requests event magnitude estimation (a). GAIA Agent will then analyze the waveform data and apply the selected magnitude estimation methods (b--c).